Get a free AI visibility report
.png)
A client forwards you a screenshot. They asked ChatGPT for the best options in their category, and three competitors came up. They did not. The message underneath says "can you fix this?" Most agencies have received some version of that email in 2026, and the honest answer is "yes, but not with the SEO retainer you are paying for now." Answering it well takes a dedicated AI search visibility package and a platform that can run it across every client, which is the job Cognizo was built for: an AEO platform with a partner program made for agencies.
This guide lays out what an AI search visibility package should actually contain: the eight services that belong on the menu, how agencies group them into tiers, what the market currently charges, and what a client should see in the monthly report. It draws on First Page Sage's October 2026 analysis of 1,146 GEO and AEO price quotes from 214 agencies, G2's March 2026 survey of 1,076 B2B buyers, and traffic data from Adobe, Pew Research Center and two academic teams. It also shows how Cognizo covers each service, since running this menu across many clients is mostly a tooling problem.
The demand is coming from buyer behavior, not from marketing trends. In G2's 2026 AI Search Insight Report, 51% of B2B software buyers said they now start research with an AI chatbot more often than with Google. 54% named AI chatbots as a source that shapes which vendors make their shortlist, the top answer in the survey, ahead of software review sites at 43%. And 69% said a chatbot led them to pick a different vendor than the one they first had in mind.
Retail shows the same shift from a different angle. Adobe's analysis of more than 1 trillion visits to US retail sites found that AI-referred traffic grew 138% year over year in May 2026, and that those visitors converted 54% better than non-AI visitors, according to Digital Commerce 360. A year earlier, AI visitors converted at about half the non-AI rate.
Two pieces of data keep the pitch honest. First, AI answers send fewer clicks. Pew Research Center found that Google users clicked a result on 8% of visits when an AI summary appeared, against 15% without one, and clicked a link inside the summary on just 1% of visits. We dig into what that means for traffic in our guide to zero-click search. Second, not every study finds AI traffic converting better. A working paper by Maximilian Kaiser and Christian Schulze, covering 12 months of data from 973 ecommerce sites, found that ChatGPT referrals converted below Google's paid and organic search, the University of Hamburg reports.
Put together, the case for a package is not "AI will flood you with traffic." It is "buyers are forming shortlists inside AI answers, and if you are not in the answer you are often not on the list." That framing matters, because it decides what you sell and how you report on it.
Before building the menu, agree with the client on the outcome. In practice, an AI search visibility package delivers four things:
Every service below exists to move one of those four. If a proposed deliverable does not, it belongs in a different retainer. Cognizo measures all four in one place, which is why the sections below show which part of the platform handles each service.
This is where every engagement starts and, for many agencies, how it is sold. The audit answers four questions: which buyer prompts matter, how often the brand appears in answers to them, which competitors appear instead, and which sources the engines cite.
A useful audit includes a defined prompt set (category, comparison, problem and branded prompts), results from each engine the client cares about, a competitor share of voice table, the top cited domains and pages, and a short list of the biggest gaps. It should be repeatable, because the baseline is what every later report is measured against.
First Page Sage found that 8% of agencies sell the audit as a standalone product, with a median price of $4,300 and a range of $1,500 to $12,000. Many others use it as the opening move of a retainer. In Cognizo, agencies can create prospect brands and run a custom AEO audit before a client signs, so the pitch deck is built on the prospect's real data instead of generic statistics. The audit pulls visibility, share of voice and cited sources from Answer Engine Insights, the same data the client will see every month after signing. Cognizo also offers a free AEO audit for brands that want a first look.
The prompt set is the client's new keyword list, and it needs the same care. Good prompt research starts from what buyers actually ask: sales call notes, support tickets, review site language, search data, and prompt volume estimates. Our guide on finding what customers ask AI walks through the sources, and Cognizo's Prompt Volumes module adds volume estimates, prompt exploration and competitive gap analysis so the agency can rank prompts by demand instead of guessing.
Tracking then has to be continuous. AI answers change constantly, and a monthly snapshot can be badly out of date by the time it reaches the client. In one publisher's Bing data, daily Copilot citations fell 97% in about two months, as we covered in our Copilot citation sources study. Cognizo tracks every prompt daily across the major answer engines, with unlimited regions, languages and competitors on every plan, which matters when an agency serves clients in several markets.
What the client gets: the prompt set, grouped by funnel stage and topic, plus daily tracking with a monthly view of movement.
AI engines cannot cite what their crawlers cannot reach or read. This module covers access for AI crawlers in robots.txt and at the CDN level, server-side rendering for pages whose main content depends on JavaScript, clean HTML structure, structured data, page speed, and internal linking to the pages that should be cited.
A page that renders its main content only in the browser, or a CDN rule that blocks AI user agents by default, can keep a client out of answers no matter how good the content is. These are fixable problems, and they are often the fastest early win in an engagement. Our guides on optimizing a site for AI crawlers and schema markup for AI search cover the checks. Cognizo's technical site audits flag these issues per domain, and AI crawler analytics show which bots actually visit and which pages they request.
First Page Sage reports that technical audits and reporting are "nearly universal" in AEO packages from the Starter tier upward. If a proposal leaves this out, ask why.
This is the largest line item in most retainers. The work is not "more blog posts." It is pages built to be quoted: a direct answer in the first lines, named authors, specific numbers with sources, comparison tables where buyers compare, and one clear question per page. Existing pages that already rank or get cited need refreshes, because freshness and accuracy affect whether they keep being chosen.
A typical month includes a mix of new pages aimed at prompts where the client is missing, refreshes of pages that are losing citations, FAQ and comparison content, and product or category page improvements for ecommerce clients. First Page Sage counts an average of 3.4 deliverables a month at the Starter tier, rising to 6.5 at Growth, 9.8 at Advanced and 14.2 at Enterprise.
This is where Cognizo saves an agency the most hours. Its Content Optimization module drafts AI-optimized articles from the gaps it finds, and its content agent can publish them automatically, so the agency team spends its time on strategy and editing rather than first drafts. The Growth plan includes 8 AI-optimized articles a month and Pro includes 20, which covers the content side of the Starter and Growth deliverable counts above for a single client.
AI engines lean heavily on third-party sources: review sites, editorial roundups, forums, video and reference pages. For many prompts, the client's own site will never be the most cited page, so the package has to include work on the pages that are.
This module starts with a source gap list: domains and pages that the engines cite for the client's prompts, that mention competitors, and that do not mention the client. From there the work is digital PR, review site programs, contributed articles, partnerships, and community participation where it is genuine. Platform behavior differs a lot. Copilot rarely cites Reddit, for example, while other engines cite it often, so the plan should follow the data for each engine rather than a generic "get on Reddit" rule. See our research on third-party AI citations and Reddit in AI answers.
First Page Sage found that digital PR appears in 33% of Growth-tier packages and 71% of Advanced ones, and original research rises from 29% to 64% across the same tiers. That jump is one of the clearest signals of what separates a mid-market package from a premium one. Cognizo's citation data and outreach targets give the team the source gap list directly, per prompt and per engine, and our guide to tracking competitors in AI search shows how to turn that list into an outreach plan.
Being mentioned is only half the job. AI answers regularly describe brands with old pricing, missing products or the wrong category, and some answers carry a negative tone pulled from an old review or forum thread. We have written about why ChatGPT gets brand descriptions wrong and how to correct it.
This module monitors how each engine describes the brand, flags factual errors and negative sentiment, traces them to the sources behind them, and fixes those sources: updating the client's own pages, correcting third-party listings, and publishing clear, current facts that engines can pick up. Cognizo's sentiment analysis and brand sentiment tracking show how the description changes over time.
Clients renew when they can see results. This module sets up the measurement: AI referral traffic segmented in analytics, landing pages that AI visitors arrive on, conversions from those visits, and crawler activity. Our GA4 guide for AI traffic covers the setup.
Because click-through from AI answers is low, reporting should not rest on traffic alone. Visibility, share of voice and citation share show progress months before referral traffic does. The reporting section below lists what a monthly report should contain. Cognizo's AI Traffic Analytics combines AI referral traffic and crawler analytics with the visibility data, and agencies can share branded reports with clients in their own branding. For a real example of visibility work tied to revenue, read how Hat Club turned AI visibility into 20x revenue with Cognizo.
Ads are arriving inside AI assistants, and some clients will want paid placement next to their organic work. This is an add-on rather than a core service, but agencies that already run paid search are well placed to offer it. The work covers competitor ad monitoring, targeting opportunities and creative. Our comparison of organic AEO and ChatGPT ads explains when each makes sense. Cognizo's AI Ads module includes an ad library, competitor tracking and targeting opportunities, so paid and organic work can be planned from the same prompt data.
Most agencies sell AI search visibility in three or four tiers, and First Page Sage's data gives a clear picture of how the market structures them. Its study normalized 1,146 quotes collected from July 6 to October 2, 2026, from pricing pages, proposals and signed contracts:
| Tier | Typical client | Monthly range | Median | Minimum term | Deliverables per month |
|---|---|---|---|---|---|
| Starter | Local and small businesses | $1,800 to $4,200 | $2,950 | 3 months | 3.4 |
| Growth | Mid-market companies | $4,500 to $11,000 | $7,400 | 6 months | 6.5 |
| Advanced | Upper mid-market | $11,000 to $22,000 | $14,600 | 6 months | 9.8 |
| Enterprise | Large enterprises | $22,000 to $60,000+ | $31,500 | 12 months | 14.2 |
Source: First Page Sage, GEO and AEO pricing statistics, October 2026.
Here is how the eight services typically map onto those tiers. This mapping is our own synthesis of the First Page Sage findings and common agency proposals, so treat it as a starting point:
| Service | Starter | Growth | Advanced | Enterprise |
|---|---|---|---|---|
| Baseline audit | Yes | Yes | Yes | Yes, multi-market |
| Prompt research and tracking | Core prompts, 1 market | Full prompt set | Full set plus competitors | Multi-brand, multi-market |
| Technical AI readiness | Yes | Yes | Yes | Yes, ongoing |
| Content creation and refreshes | A few pages a month | Steady monthly program | Larger program plus original research | Full program, several teams |
| Off-site citation building | Light | Review sites and outreach | Digital PR | Digital PR at scale |
| Brand accuracy and sentiment | Monitoring | Monitoring and fixes | Full program | Full program, all markets |
| Attribution and reporting | Monthly report | Monthly report and call | Monthly plus quarterly review | Custom dashboards |
| AI ads | No | Optional | Optional | Optional |
Two things stand out in the data. Multi-market or multilingual coverage appears in 67% of Enterprise packages against 24% at Advanced, so international work is the main step from Advanced to Enterprise. And the minimum term doubles from three to six months once a client moves past Starter, which reflects how long citation building and content take to show results.
Retainers dominate, but they are not the only model. First Page Sage's breakdown:
| Model | Share of agencies | Typical range | Median |
|---|---|---|---|
| Monthly retainer | Most agencies | $1,800 to $60,000+ | $6,850 |
| Project or sprint (30 to 90 days) | 13% | $5,000 to $40,000 | $14,500 |
| One-time audit (2 to 6 weeks) | 8% | $1,500 to $12,000 | $4,300 |
| Hourly consulting | 6% | $125 to $325 an hour | $205 an hour |
Three findings are useful when setting your own prices:
The premium only holds if the package contains work that SEO retainers do not: prompt-level tracking across engines, source gap analysis, citation building and AI-specific reporting. A relabeled SEO package at an AEO price will not survive the first quarterly review.
On the cost side, tooling is usually the line agencies underestimate. Running separate subscriptions for every client gets expensive quickly. Cognizo's agency partner program offers bulk pricing and consolidated billing, so one bill covers the whole client portfolio, and a single agency workspace with role-based access holds every client and prospect. Because Cognizo also produces the content and runs the technical audits, one platform replaces the separate tracking, content and audit subscriptions that eat into a retainer's margin.
The report is where the package proves its value. It should read like a work queue: what changed, why, and what happens next. A solid monthly report covers:
| Section | What it shows | Why it matters |
|---|---|---|
| AI visibility | Share of tracked prompts where the brand appears, by engine | The core presence metric |
| Share of voice | Brand mentions against named competitors, see how to measure AI share of voice | Shows competitive position |
| Citation share | How often the client's pages are cited, owned vs earned | Shows source authority |
| Wins and losses | Prompts gained or lost this month, and to whom | Turns data into actions |
| Source gaps | Cited domains that mention competitors but not the client | Feeds outreach and PR |
| Brand accuracy and sentiment | Errors and tone changes in how engines describe the brand | Protects reputation |
| AI referral traffic and conversions | Visits and conversions from AI sources | Ties work to revenue |
| Work completed and next actions | Pages shipped, fixes made, the plan for next month | Shows the agency is moving |
Our guides to AI visibility metrics and building an AI visibility report for clients go deeper on each metric. In Cognizo, these metrics come straight from the platform, and agencies can export them or share reports in their own branding. Teams that prefer to build custom dashboards can pull the same data through Cognizo's API and MCP server, so an account manager can ask "which prompts did this client lose this month" from Claude and get the answer without opening a dashboard.
Some deliverables look good in a proposal and do not move any of the four outcomes. Leave these out:
Delivering eight services across ten or fifty clients is not possible with spreadsheets and manual prompt checks. When choosing a platform, agencies should look for:
Cognizo was built to cover that full list in one place. Here is how each service on the menu maps to the platform:
| Service | Where it runs in Cognizo |
|---|---|
| Baseline audit | Prospect brands with custom AEO audits, built on Answer Engine Insights |
| Prompt research and tracking | Prompt Volumes for research, daily tracking with unlimited regions, languages and competitors |
| Technical AI readiness | Technical site audits and AI crawler analytics |
| Content creation and refreshes | Content Optimization, content agent with auto-publishing, content refreshes |
| Off-site citation building | Citation share, source and page type classification, outreach targets |
| Brand accuracy and sentiment | Sentiment analysis and mention share |
| Attribution and reporting | AI Traffic Analytics, exports, branded reports, API and MCP |
| AI ads | AI Ads module with ad library and competitor tracking |
On top of the platform, the agency program adds prospect audits, bulk pricing and consolidated billing, branded reporting, a listing in the Cognizo Agency Marketplace for inbound leads, a dedicated agency partner for go-to-market support, AEO education for the team, and co-selling and co-marketing opportunities. The difference from tracking-only tools is that the same workspace that finds a gap also drafts the page, audits the site and reports the result, so the agency is not stitching four products together for every client. Our roundup of AI visibility tools for agencies and our comparison of AI visibility platforms cover the alternatives.
The strongest pitch opens with the prospect's own answers, not industry statistics. Run their top ten or twenty buyer prompts across the engines before the meeting, and open with what the buyer sees today: who appears, who does not, and which sources the engines quote. A marketing director who watches a competitor named three times in answers to their most important question rarely needs convincing that the problem is real.
From there, a proposal that converts usually has five parts:
Keep the expectations honest. Promise measured progress on a defined prompt set, not a position in ChatGPT, and explain that AI answers move daily, so the report will show swings as well as gains. Clients who understand that from the first call stay through the slow months, and clients who were promised a guaranteed ranking leave at the first bad week.
A prospect brand in Cognizo makes this practical at scale. The agency can set up the prospect's prompts and competitors, run the audit, and present real data in the first meeting. Keep that audit, because the baseline in the proposal should be the same baseline the first monthly report is measured against.
A clear first quarter keeps the client patient while results build. A common structure:
Days 1 to 30: baseline and fixes. Agree the prompt set and competitors, run the baseline audit across the client's engines, fix technical blockers, correct the most visible brand description errors, and set up AI traffic segments in analytics. Deliver the baseline report. In Cognizo, this month runs on the baseline audit and a technical site audit.
Days 31 to 60: content and sources. Publish the first answer-ready pages for the highest-value missing prompts, refresh existing pages that are close to being cited, and start outreach to the top domains on the source gap list. Report the first movement in visibility and citations. Cognizo's content agent drafts the first pages from the gaps, and its outreach targets become the PR list.
Days 61 to 90: scale what works. Double down on the page types and sources that produced wins, expand the prompt set if needed, and present a quarterly review that compares results with the baseline and sets the plan for the next quarter. The branded report in Cognizo shows the client the movement since day one.
Use a system like Cognizo that tracks the client's prompts daily, so each of these checkpoints rests on fresh data rather than a one-off check.
At minimum: a baseline audit, prompt research with ongoing tracking, technical AI readiness, answer-ready content and refreshes, off-site citation building, brand accuracy and sentiment monitoring, and AI traffic attribution with monthly reporting. AI ads can be added as an option. Each service should move presence in answers, accuracy of the brand description, citations, or measured results.
First Page Sage's analysis of 1,146 quotes from 214 agencies found a median retainer of $6,850 a month in Q3 2026. Medians by tier were $2,950 for Starter, $7,400 for Growth, $14,600 for Advanced and $31,500 for Enterprise. One-time audits had a median of $4,300 and sprints a median of $14,500.
It overlaps but is not the same. Technical health and good content help both. AI visibility adds prompt-level tracking across several engines, analysis of which sources each engine cites, citation building on third-party sites, and monitoring how engines describe the brand. That extra work is why AEO retainers carry about a 45% premium over SEO retainers in First Page Sage's data. Our GEO vs traditional SEO comparison goes through the differences in detail.
Technical fixes and brand description corrections can show within weeks. Content and citation building usually take a few months, which is why minimum terms move from three to six months beyond the Starter tier. Daily tracking shows early movement on individual prompts before the overall numbers shift.
Track visibility, share of voice and citation share on the agreed prompt set, then AI referral traffic and conversions in analytics. Because users click links inside AI answers rarely, as low as 1% in Pew's data for Google's AI summaries, visibility and shortlist influence matter more than raw clicks. For B2B clients, ask new leads how they found the company.
The engines the client's buyers use. For most clients that means ChatGPT, Google AI Overviews and AI Mode, Gemini and Perplexity, plus Copilot where buyers work inside Microsoft tools. Each engine cites different sources, so tracking them separately is part of the job.
It works as an entry point. 13% of agencies in First Page Sage's data sell sprints of 30 to 90 days, and 8% sell standalone audits. But AI answers keep changing, so results fade without ongoing tracking and content. Most agencies use the audit or sprint to start a retainer.
Look for multi-client management, daily prompt-level tracking across engines, citation source data, content execution, technical audits, branded reporting and API access in one place. Cognizo covers that list in one workspace, from tracking to content to reporting, and adds an agency program with prospect audits, bulk pricing, consolidated billing and marketplace listings for inbound leads.
.svg)
.svg)
